EnergyPlus Model Research Articles

A hybrid load prediction method of office buildings based on physical simulation database and LightGBM algorithm

Building load prediction plays an important role in building energy savings and mechanical and electrical system optimization control. The dynamic energy consumption can be accurately calculated using the traditional physical energy simulation method that entails a complex setup and verification process due to the numerous input parameters required. It is also difficult to change the physical model once it has been determined. The data-mining method is fast in calculation and simple to use, but its prediction accuracy is limited by historical data quality. It is difficult to predict the load of new buildings without historical data. To solve these problems, this study proposes a hybrid building load prediction method for office buildings. The proposed method uses EnergyPlus to generate a building load database that includes than 25.14 million data cases, covering 35 types of building geometry and 2870 building examples. Based on the above database, the LightGBM algorithm was selected to extract feature variables that affect the load and build a load prediction model. The training results show that there are 24 key feature variables for office building load prediction. The hourly MAPE of the cooling load prediction model is 6.95 % and RMSE is 4.31 W/m2, and the hourly MAPE of the heating load prediction model is 7.09 % and RMSE is 11.64 W/m2 compared with EnergyPlus model. Two actual office buildings are selected as case studies to validate the model prediction accuracy. Results show that comparing predicted results with measured data, the hourly cooling load of the MAPE is 12.42 %. Coparing predicted results with actual heating load, daily MAPE is 7.97 %.

Large-scale energy cost optimization and performance analysis for dedicated outdoor air system: simulation results from ASHRAE RP-1865

This is the first journal article from the ASHRAE research project RP 1865, “Optimizing Supply Air Temperature Control for Dedicated Outdoor Air Systems,” focusing on the large-scale energy performance evaluation of the optimal control of dedicated outdoor air systems (DOAS). DOAS is a type of air handling system that is installed in buildings to ensure proper ventilation for the built environment. It is acknowledged that DOAS with proper operation strategies can provide sufficient ventilation and dehumidification while achieving energy efficiency. In this regard, there have been efforts to develop advanced control sequences (e.g., optimal control) for DOAS. Still, limited studies have demonstrated the effectiveness of such DOAS controls on a larger scale under varying climate zones with different DOAS configurations. To fill the research gap, this study aims to quantify the energy-saving potentials of applying an optimal supply air temperature (SAT) control strategy for DOAS considering diverse configurations of HVAC systems in various climate conditions across the United States. For this study, we first developed a total of 180 EnergyPlus models accommodating different types of buildings (i.e., detailed medium-sized office and primary school) and HVAC systems (i.e., DOAS types, DOAS sizes, and terminal units) under varying ASHRAE climate zones (i.e., 2A, 3B, 4A, 4C, 6A, and 6B). Next, genetic algorithm (GA)-based optimizations were conducted to determine the optimal DOAS SAT control that would minimize the energy cost of the entire HVAC system operation. Our study of comparing the optimal control to a typical rule-based control (i.e., outdoor air temperature-based reset control) revealed approximately 4–34% of energy cost saving potential by optimizing the DOAS SAT control sequence.

Model-predictive space heating control for energy flexibility – a case study using a long short-term memory neural network surrogate model and a genetic optimization algorithm

This paper presents a case study where a model predictive control (MPC) logic is developed for energy flexible operation of a space heating system in an educational building. A Long Short-Term Memory Neural Network (LSTM) surrogate model is trained on the output of an EnergyPlus building simulation model. This LSTM model is used within an MPC framework where a genetic algorithm is used to optimize setpoint sequences. The EnergyPlus model is used to validate the performance of the control logic. The MPC approach leads to a substantial reduction in energy consumption (7%) and energy costs (13%) with improved comfort performance. Additional energy costs savings are possible (7–16%) if a sacrifice in indoor thermal comfort is accepted. The presented method is useful for developing MPC systems in the design stages where measured data is typically not available. Additionally, this study illustrates that LSTM models are promising for MPC for buildings.

ASSESSMENT OF THE LEVEL OF ENERGY EFFICIENCY OF A SCHOOL TAKING INTO ACCOUNT THE USE OF LOCAL RENEWABLE ENERGY SOURCES

Energy efficiency is one of the key topics that have become very relevant in recent years. In buildings, as the final consumers of energy, the greatest potential for energy saving is concentrated. And at the same time, an important component of energy efficiency is the optimal use of energy resources, where the lion's share falls on the consumption of thermal energy. In this context, heating schedules become an important tool for achieving this goal. In this article, the authors investigate the energy efficiency of different heating and ventilation modes for a school located in two climatic zones of Ukraine: I (Kyiv) and II (Odesa). Using the Design-Builder software, which is based on the EnergyPlus model, the authors model five intermittent modes for the heating and ventilation system. The authors also compare the energy efficiency of different heat sources: a boiler on biofuel, a heat pump and a solar panel. It was found that the introduction of heating modes according to the schedules allows to reduce energy consumption for heating by 10 % to 20 % depending on the mode and climatic zone, and the use of renewable energy sources based on the introduction of intermittent heating and ventilation modes for the school shows efficiency at 79 %-80 %. The authors also analyze the impact of thermomodernization of enclosing structures on the energy efficiency of buildings. The conclusion on the introduction of intermittent heating modes for non-insulated buildings is an extremely effective measure for energy saving, and the impact of the measure on energy saving of the building will be noticeable even after thermal insulation. The implementation of all improvements allows reducing energy consumption for heating by 90 %.

Prediction models for building thermal mass of intermittently heated rooms for balancing energy consumption and indoor thermal comfort

Building thermal mass (BTM) plays an essential role in maintaining a comfortable indoor environment for intermittently heated rooms (IHR). However, how to properly determine the BTM in IHR is unclear for the comprehensive consideration of the opposite effects on the reduction of heating energy consumption (HEC) as well as the improvement of indoor thermal comfort (ITC). This study aims to investigate the comprehensive impact of BTM on the HEC and ITC of IHR in the hot summer and cold winter (HSCW) zone of China using the validated EnergyPlus model. The results indicate that BTM can significantly improve the ITC of IHR, whereas increase HEC. The degree of impact is mainly related to the internal wall heat accumulation coefficient, external wall insulation layer thermal resistance, and outdoor air mean temperature during non-heating periods. Response surface methodology is further employed to quantify the effects of BTM on HEC and ITC of IHR. Based on the predictive models, the range of parameter variations for BTM that simultaneously meet HEC and ITC requirements is obtained. The predictive models can provide a theoretical basis for balancing HEC and ITC in intermittently heated buildings in the HSCW zone of China.

Integrating daylight with solar thermal modeling: A dynamic calculation method for complex fenestration systems

This study developed a mathematical model named the DL-E model, which links annual dynamic daylight and solar heat gain calculations for Complex Fenestration Systems (CFSs). The DL-E model utilizes the Klems Bidirectional Scattering Distribution Function (BSDF) to characterize CFSs and can consider the shading effect of obstacles, as well as solar radiation reflected from obstacles and the ground. The DL-E model uses the Perez all-weather model to calculate transmitted solar radiation contributed by diffuse solar radiation, rather than the simplified sky model used in EnergyPlus. Results calculated with the DL-E were compared with those computed with EnergyPlus, and it was found that for most of the results, the both methods yielded closely aligned outcomes. However, for the diffuse transmitted solar radiations, reflected solar radiation from the ground in the case with a vertical obstacle, and the reflected solar radiations from the obstacle, the values calculated with DL-E showed significant differences compared to those calculated with EnergyPlus due to the use of the simplified sky model in EnergyPlus. In contrast, the values calculated with the DL-E model which used the Perez all weather sky model displayed very small difference compared to those calculated with Radiance, which means the values derived with EnergyPlus may produce significant errors. Additionally, a new workflow has been developed for calculating building load and energy for CFSs to avoid the manual conversion of the BSDF files.

Unveiling overlooked aspects of model predictive control for building air conditioning systems

The ongoing research on model predictive control (MPC) for building air conditioning systems predominantly centers on improving the predictive capabilities of system models. In this paper, the impacts of three additional pivotal factors on MPC performance are assessed by examining a generic MPC design for a typical variable air volume (VAV) system that serves large commercial buildings. The three factors encompass the nuanced reformulation of optimization, the judicious relaxation of constraints, and the meticulous tuning of parameters. Detailed case studies with an integrated Modelica and EnergyPlus model of the US Department of Energy's Commercial Reference Building are conducted. The results confirm that the optimization formulation, along with relaxation methods, significantly affects MPC performance in terms of energy savings, zonal thermal comfort level, and computational demand. They also reveal that the impact of the MPC control parameters on the energy savings and thermal comfort may vary by season and can be non-monotonic.

Energy demand parametric analysis and geothermal heat exchanger design applied to a nearly zero energy PV building in northern Italy

Nearly-Zero-Energy and Zero-Emission Buildings are one fundamental part of the world strategies addressed at mitigating the global warming trend and coping with the goal of setting 1.5 °C the planet temperature increase with respect to the pre-industrial conditions. The present paper refers to a specific case study, the Smart Energy Building (SEB) located in the Savona Campus of the University of Genoa. The SEB is a very innovative building for both the envelope (ventilated highly insulated facades) and the energy systems (including a ground coupled heat pump and an exhaust air-to-air heat pump for building ventilation); the building is equipped also with a PV module field on the rooftop for electricity production. The present paper first evaluates the heating and cooling loads of the building by means of an EnergyPlus model and analyses the impact of different control strategies related to the temperature setpoints. In particular, case #1 refers to a minimum temperature to be assured by heating only in winter and to a maximum one to be avoided by cooling only during summer; on the other hand, in case #2 the temperature inside the building is controlled within a defined range (both heating and cooling modes) all year long. From EnergyPlus simulations, the hourly heat loads have been averaged as monthly values and employed for the calculation of the required overall length of the ground heat exchangers of the geothermal heat pump of the SEB building. The ground-side analysis is performed by applying the Authors’ ASHRAE-Tp8 method, in its recently released web version. For the BHE (borehole heat exchanger) calculations, a parametric analysis has been performed in terms of ground thermal conductivities. Finally, the PV electrical production has been estimated by EnergyPlus simulations and compared with the corresponding measured one, thus showing the net energy zero behaviour of the present building.

Application of transfer learning to overcome data imbalance and extrapolation for model predictive control: A real-life case

This paper proposes a transfer learning (TL)-based control-oriented model development framework. In particular, this study examines the transferability from virtual (source) to existing (target) buildings to overcome the data imbalance issue of the data-driven approach. The target system is a cooling system comprising two supply air fans and four condensing units. First, synthetic data rich enough to provide fundamental knowledge about the target system were generated using the EnergyPlus model. A data-driven model was subsequently developed to learn the underlying dynamics of the system. By adopting TL using an imbalanced dataset measured from the target system, the knowledge that the model learned from the virtual data was transferred to the target system of the existing building. The results showed that the transfer learning model could accurately describe the dynamic behavior of the target system and predict the supply air temperature with marginal errors (CVRMSE: 5.4%, MAE: 0.96 ℃). In other words, the TL from virtual to existing buildings can overcome the data imbalance issue for developing a reliable data-driven model.

Preparation study and simulation of high transmittance W–Sn–Zr Co-doped VO2 thin films prepared by sol-gel method for energy-saving smart windows

The thermochromic properties of vanadium dioxide (VO2) hold great promise for applications in smart glass windows. VO2 films exhibit a high phase transition temperature (Tc), low visible transmission (Tlum), and low solar modulating ability (ΔTsol), making them inadequate for the requirements of smart glass windows. While some improvement in VO2 film performance can be achieved through single-element doping, these limited enhancements still fall short of ideal results. Therefore, there is a pressing need for the improvement of VO2 films. In this study, we utilized the sol-gel method to prepare VO2 samples simultaneously doped with three elements: W, Sn, and Zr. Microscopic characterization confirmed the successful incorporation of these elements. Optical measurements, including ultraviolet–visible–near-infrared spectroscopy, were conducted to assess the optical properties of the films. Ultimately, outstanding Tlum (69%) and ΔTsol (14%) were achieved, along with an appropriate Tc (27.6 °C). Energy consumption simulations were performed using the EnergyPlus model with experimental data, revealing that compared to conventional glass, VO2 glass reduces total cooling energy consumption by 9.8%. In comparison to previous work, the multi-element-doped films push the performance limits of VO2 films, offering a viable approach for fabricating superior VO2 smart glass windows.

Infiltration Models in EnergyPlus: Empirical Assessment for a Case Study in a Seven-Story Building

The current decarbonization transition to be achieved by 2050 according to the European Council has given great prominence to the use of Digital Twins as tools for energy management. For their correct operation, it is essential to control the uncertainties of the energy models, which lead to differences between the measured and predicted data. One of the key parameters that is most difficult to assess numerically is air leakage. The existent infiltration models available in EnergyPlus were developed to be applied in low-rise residential buildings with fewer than three stories. Therefore, it is common to rely on air leakage equations employing predefined coefficients. This research presents an empirical assessment of the performance of two EnergyPlus air leakage models, the “Effective Leakage Area” and the “Flow Coefficient”, in predicting dynamic infiltration within the attic of a seven-story building. Blower door tests, along with the application of CO2 tracer gas, were conducted to establish coefficients for the models. Then, they were evaluated in three independent periods according to the criteria established in the American Society for Testing Material D5157 Standard. Those models that only used in situ coefficients consistently met the standard across all three periods, demonstrating for both equations their accurate performance and reliability. For the best model derived from tracer gas data, the R2 and NMSE values are 0.94 and 0.019, respectively. In contrast, the model developed using blower door test data and EnergyPlus default values presented a 64% reduction in accuracy compared to the best one. This discrepancy could potentially lead to misleading energy estimates. Although other software options exist for estimating infiltration, this study specifically targets EnergyPlus users. Therefore, these findings offer valuable insights to make more informed decisions when implementing the infiltration models into energy simulations for high-rise buildings using EnergyPlus.

Parametric analysis of phase change materials within cold climate buildings: Effects of implementation location and properties

Integrating phase change materials (PCMs) into buildings is a proven method to reduce heating and cooling loads in alignment with international targets to reduce energy and greenhouse gas emissions. However, although promising, PCM uptake has been limited, particularly in cold climates due to a lack of knowledge about their performance and lack of optimized PCM solutions. The objective of this study was to provide a comprehensive analysis of PCMs, integration strategies, and properties that lead to the largest space conditioning reductions in four cold climate cities within Canada. An experimentally validated EnergyPlus model was developed and used to simulate a thin layer of PCMs on surfaces adjacent to a conditioned space to predict their effects and trends that occur due to their placement in various locations within a room, geographic locations, and melting temperatures. It was found that the most effective location for PCM installation was the room walls for heating load reductions and floor for cooling load reductions, due to the incident solar energy profiles in each conditioning season. In the warmer two cities studied, the greatest benefits per unit PCM occurred with PCM in the floor, while the cooler two cities had the greatest benefits per unit PCM with it installed in the walls. The optimal choice of PCM to decrease heating load was 1–2 °C above the heating setpoint, while for decreasing cooling loads it was 1–2 °C below the cooling setpoint, regardless of the choice of conditioning setpoints or geographic location.

Impact of building porosity on exterior convective heat transfer coefficients: An experimental and computational parametric study

Exterior convective heat transfer is vital for connecting a building and its surrounding environment. With an increasing focus on porous buildings for sustainable design, accurately calculating exterior convective heat transfer is critical, and also poses significant challenges. This study used experimental and computational fluid dynamics (CFD) techniques to assess the impact of building porosity on exterior CHTCs. Onsite measurements at an actual porous building were conducted first for validation, which was further extrapolated by linking the actual porous building with idealized porous buildings. Parametric CFD simulations were then performed on hypothetical isolated buildings with ten different idealized geometries (including one fully enclosed building as a baseline and nine porous buildings) and two wind directions (perpendicular and oblique). The resulting exterior CHTCs obtained through CFD simulations were compared with those from exterior CHTC models in EnergyPlus, along with the resulting differences in cooling load. The findings revealed that the maximum difference in exterior CHTCs between CFD simulations and exterior CHTC models in EnergyPlus was nearly threefold, leading to cooling load deviations of up to 19%. From a practical perspective, this study highlights the impact of building porosity on exterior convective heat transfer, underscoring the importance of considering building porosity in building energy simulation for more accurate cooling load prediction.

Sensitivity analysis of an automated fault detection algorithm for residential air-conditioning systems

The state of the art of fault detection and diagnosis (FDD) for residential air-conditioning systems is expensive and not yet amenable to widespread implementation. FDD for homes can significantly reduce utility costs, and increase the lifespan of the equipment. The cost barriers currently, however, make FDD for homes economically unviable for large scale implementation. In prior work, we offered a solution to reduce FDD costs by proposing an automated fault detection algorithm to serve as a screening step before more expensive FDD tests can be conducted. The algorithm uses only the home thermostat and local weather information to identify thermodynamic parameters and detect high-impact air-conditioning faults, including those that occur during equipment installation. We had tested the algorithm on a single EnergyPlusTM model of a home in Orlando, Florida. The thermodynamic parameter identification process is highly nonconvex involving several local optimal solutions. In this paper we propose a novel method to select the best model for fault detection from among the list of local optimal solutions to make the algorithm more robust to homes of different construction, without which the fault detection process would be infeasible. Another unique contribution of the paper is implementing the solution on real-world data. We also bring the algorithm closer to market by testing it on real-world data. We implement the algorithm on data obtained from experiments conducted by the Florida Solar Energy Center (FSEC) on a laboratory home equipped with a heat pump where faults were intentionally added for a period of seven months. The algorithm successfully detected an undercharge fault with 70.6% accuracy, concurrent duct leakage and undercharge faults with 85.2% accuracy, and duct leakage faults with 69.1% accuracy. A sensitivity analysis is also performed on EnergyPlus models of nine types of homes that vary in construction to demonstrate the robustness of the algorithm. The algorithm achieves an average accuracy of 71% for no-fault condition, 77% for 40% undercharge fault, and 76% for duct-leak fault.

Energyplus model of double skin façade and diffuse ceiling ventilation

This paper introduces an Energyplus model to combine the double skin façade and diffuse ceiling ventilation systems with weather-adjusted dynamic control for both heating and cooling purposes. The proposed Energyplus model is validated with measurement data for both the heating mode and cooling mode. The double skin façade, diffuse ceiling, and the room are modeled as different thermal zones in Energyplus. The ventilation between the adjacent zones is controlled by zone ventilation and zone mixing. The cooling mode of the double skin façade is modeled by wind and stack. An energy management system (EMS) controls the ventilation modes based on outdoor conditions and indoor air temperature. The geometry of the model is identical to the experiment room in the façade lab facing south, with the same internal load and ventilation mode. The model results are then compared to the experimental data, including both heating mode and cooling mode for 4 thermal zones: The DSF zone, the DSF top zone, the diffuse ceiling zone, and the room zone. For heating mode, the average discrepancy of all 4 thermal zones between the model results and experimental data is 13,75%, while for cooling mode, the average discrepancy is 13.78%. The model performance of the room zone is the highest, which has a discrepancy of 5.31% for the heating mode and 3.67% for the cooling mode compared to the experimental data.

Economic Applicability of Solar Tracking Photovoltaic Systems in Commercial Buildings: Case Study in South Korean Climate

This study investigated the applicability of a tracking photovoltaic (PV) system installed in the roof area of a commercial building. Because PVWatts is the only PV module with a tracking feature in EnergyPlus, its electricity generation was validated through comparisons with detailed PV modules in EnergyPlus. The tracking PV system generated 26.8–35.5% more electricity annually than a fixed system in the climate of Incheon (S. Korea). The load coverage analysis of the tracking PV system was conducted with the reference commercial building model in EnergyPlus. Approximately 14% of the total building electric demand, including heating, ventilation, and air-conditioning; lighting; and equipment, was met by one PV array. Finally, the life cycle cost analysis of the tracking PV system was conducted by considering the net present value, which includes the initial installation and operation costs. The initial investment was returned after approximately 8 years, assuming between two and six tracking PV arrays were installed. Moreover, up to 26.8% cost savings were achieved in 15 years compared to the case without any PV arrays.

Urban Building Energy Modeling with Parameterized Geometry and Detailed Thermal Zones for Complex Building Types

Urban building energy modeling (UBEM) has attracted wide attention to the requirement for global carbon emission reduction. This paper presents a UBEM tool, AutoBPS-Param, to generate building energy models (BEMs) with parameterized geometry and detailed thermal zones, especially for complex building types, considering the shading effect from surrounding buildings simultaneously. Three building number scales and four scenarios were analyzed in the hotel-related buildings in Changsha, China. For the prototype modeling of Scenario 1, eighteen prototype building energy models for six building types in three vintages were created, and their simulation results were aggregated based on their representative floor areas. For AutoBPS-Param of Scenario 4, the method created one EnergyPlus (Version: 9.3.0) model for each building. The geometry of the prototype model was scaled and modified based on the target building’s length, width, and number of stories. The surrounding buildings were also added to the AutoBPS-Param simulation to better capture the urban dynamic impact. The results showed that the annual electricity and natural gas energy use intensity (EUI) of the pre-2005 HotelOffice prototype model was 172.25 and 140.45 kWh/m2. In contrast, with the AutoBPS-Param method, the annual electricity EUIs of 71 HotelOffice buildings constructed before 2005 ranged from 159.51 to 213.58 kWh/m2 with an average of 173.14 kWh/m2, and the annual gas EUIs ranged from 68.02 to 229.12 kWh/m2 with an average of 108.89 kWh/m2. The proposed method can better capture the diversity of urban building energy consumption.

A Comparative Study on the Distribution Models of Incident Solar Energy in Buildings with Glazing Facades

The accurate distribution of solar energy on indoor walls is the basis of simulating the indoor thermal environment, and its specific distribution changes all the time due to the influence of solar azimuth and altitude angle. By analyzing the assumptions of each model, the existing solar energy distribution models are eight kinds in all and are divided into three categories. The solar radiation models in TRNSYS, EnergyPlus, and Airpak software all use the absorption-weighted area ratio method, which assumes that a single interior surface is a whole, but the detailed assumptions of the models used in the three software are different. In the Radiosity-irradiation method, the indoor surfaces are discretized into small surfaces for calculation. The calculation accuracy of solar radiation distribution indoors can be controlled by the number of discrete small surfaces. The Radiosity-irradiation method is implemented by using Matlab software programming in this paper. Through the numerical calculation and analysis of typical cases, the solar distribution results of the absorption-weighted area ratio method and the Radiosity-irradiation method all show the asymmetry. The asymmetrical ratio of direct solar radiation varies during the time between 7.96–9.89, and the minimum turns up at 11:30 in the summer solstice. The asymmetrical ratio of diffuse solar radiation is 3.23 constantly. The asymmetrical ratio of total solar energy is mainly influenced by the direct and diffuse solar feat gain and its value changes in the range from 3.4 to 4.45 in the summer solstice. Calculation comparison and error analysis on the solar radiation models used in TRNSYS, EnergyPlus, and Airpak software are conducted. There are significant errors in the simulation results of all three software. TRNSYS has the highest error among the three software as its results do not change over time. For EnergyPlus, the distribution ratio of floor 1 is too large. Airpak has the smallest error, but the solar radiation distribution ratios of the indoor surfaces near the south glazing facade are underrated, especially the indoor surfaces that have not been exposed to direct solar radiation.

Rooftop unit comparison calculator: a framework for comparing performance of rooftop units with building energy simulation

The applications of building energy simulation (BES) in designing heating, ventilation, and air conditioning (HVAC) systems are limited by the high costs of developing simulation models and the lack of references for determining the model parameters. This paper presents a software framework for selecting designs for rooftop unit HVAC (RTU) systems with BES. Specifically, this framework reduces the cost of using BES by automating the generation of EnergyPlus models. It also employs a systematic method for determining model parameters based on well-accepted datasets. We applied this framework in a comprehensive assessment of an advanced design of RTU systems in which 478 EnergyPlus models were developed without human involvement. The assessment reveals that replacing a constant-speed fan/coil with a multiple-speed fan/coil may not guarantee better overall performance. It also suggests the benefits of replacing furnace coils with heat pumps are subject to utility cost, weather conditions, and heating load profiles.

A real-time operational carbon emission prediction method for the early design stage of residential units based on a convolutional neural network: A case study in Beijing, China

Buildings are currently responsible for 39% of global energy related carbon emissions. Thus, carbon reduction in the construction industry is crucial to global climate issues. This study contributes a real-time operational carbon emission prediction method for the early design stage of residential units based on a convolutional neural network (CNN). The CNN model was trained based on a dataset with 2000 real residential units focusing on typical cases in northern China, particularly in Beijing, which is a significant energy consumption area. The R2 values of the operational-stage heating and electricity carbon emissions in the validation dataset reached between 0.91 and 0.98. The trained model based on this method can provide the real-time prediction of the carbon emission potential of proposed residential units to buttress architects to make scientific design decisions. The model was then validated on new cases that varied in the total area, rooms, orientation, window locations, and window sizes. The performance of the CNN model is verified by comparing the results with the output of an EnergyPlus model. While the variation trends of the output results of the two methods are found to be consistent, the CNN model results still have certain deviations from those of the EnergyPlus model. Consequently, the proposed method can reflect the variation of operational-stage carbon emissions due to the change of the floor layouts of residential units, and has a faster speed and higher convenience than the simulation-based method.